Understanding Witchcraft and Sorcery in Southeast Asia 9780824840464

Table of contents :
CONTENTS
PREFACE
1 Introduction
2 The Relativity of Magical Malevolence in Urban Thailand
3 Witchcraft, Sorcery, Fortune, and Misfortune among Lisu Highlanders of Northern Thailand
4 Witches, Fortune, and Misfortune among the Shan of Northwestern Thailand
5 Anger, Anxiety, and Sorcery: An Analysis of Some Nuaulu Case Material from Seram, Eastern Indonesia
6 Social and Symbolic Aspects of the Witch among the Nage of Eastern Indonesia
7 Observations on the Practice of Sorcery in Java
8 Sorcery and the Law in Modern Indonesia
9 Knowledge, Power, and Personal Misfortune in a Malay Context
10 Return to Sender: A Muslim Discourse of Sorcery in a Relatively Egalitarian Society, the Gayo of Northern Sumatra
11 Perceptions from Within: Malign Magic in Indonesian Literature
AUTHOR INDEX
SUBJECT INDEX
CONTRIBUTORS

Citation preview

Understanding Witchcraft and Sorcery in Southeast Asia

Understanding Witchcraft and Sorcery in Southeast Asia Edited by C . W . WATSON AND ROY ELLEN

UNIVERSITY OF HAWAII PRESS

Honolulu

© 1993 University of Hawaii Press Allrightsreserved Printed in the United States of America 98 97 96 95 94 93 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data Understanding witchcraft and sorcery in Southeast Asia / edited by C. W. Watson and Roy Ellen, p. cm. Rev. papers presented at the First Canterbury International Symposium on Southeast Asian Studies, held at the University of Kent, Sept. 25-28,1989. Includes bibliographical references and index. ISBN 0-8248-1515-7 (alk. paper) I. Watson, C. W. II. Ellen, R. F., 1947. III. Canterbury International Symposium on Southeast Asian Studies (1st: 1989 : University of Kent) BF1584.A785U53 1993 133.4'3 '0959—dc20 93-5293 CIP

University of Hawaii Press books are printed on acid-free paper and meet the guidelines for permanence and durability of the Council on Library Resources

Design by Kenneth Miyamoto

CONTENTS

Preface

vii

Introduction R O Y ELLEN 2 The Relativity of Magical Malevolence in Urban Thailand 1

Louis G O L O M B

3 Witchcraft, Sorcery, Fortune, and Misfortune among Lisu Highlanders of Northern Thailand E. PAUL DURRENBERGER 4 Witches, Fortune, and Misfortune among the Shan in Northwestern Thailand N I C O L A TANNENBAUM 5 Anger, Anxiety, and Sorcery: An Analysis of Some Nuaulu Case Material from Seram, Eastern Indonesia ROY ELLEN 6 Social and Symbolic Aspects of the Witch among the Nage of Eastern Indonesia GREGORY FORTH 7 Observations on the Practice of Sorcery in Java RONNY NITIBASKARA

8 Sorcery and the Law in Modern Indonesia

1

27 47 67 81 99 123

H E R M A N SLAATS

AND KAREN P O R H E R

135

9 Knowledge, Power, and Personal Misfortune in a Malay Context M I C H A E L G . PELETZ

10 Return to Sender: A Muslim Discourse of Sorcery in a Relatively Egalitarian Society J O H N R. B O W E N 11 Perceptions from Within: Malign Magic in Indonesian Literature C . W . WATSON

Author Index Subject Index Contributors

v

149

179 191

213 217 221

PREFACE

T H E INSPIRATION for this volume was the first Canterbury International Symposium on Southeast Asian Studies held at the University of Kent between 25 and 28 September 1989, under the tide "The Manipulation of Mystical Agency and Explanations of Personal Misfortune in Southeast Asia." Ten of the papers presented at the conference have been subsequently revised for this volume. An introduction establishes the parameters of the discussion, reflects upon its themes, and, where possible, draws conclusions. The specific tasks which the symposium set itself were fivefold: to present detailed case studies of witchcraft and sorcery beliefs and practices for a cross-section of Southeast Asian peoples in societies where these have been poorly studied hitherto; to bring together different disciplinary specialists to achieve this end; to examine the bases for variation and continuity of these phenomena in the region; to see if there is anything culturally distinctive about such mystical knowledge and its manifestations in the Southeast Asian context; and to assess what contribution their study might make to our understanding of comparable phenomena globally. The extent to which this ambitious agenda was addressed will be evident from what follows, though it is our belief that the resulting volume serves to fill a significant gap in the existing ethnography.

For their generous financial support, we would like to thank the Wenner-Gren Foundation, the British Academy, and the Centre for Southeast Asian Studies at the University of Kent. Robert Barnes, David Hicks, Mark Hobart, Barry Hooker, Leo Howe, Werner Kraus,

vii

viii

Preface

and Andrew Turton chaired the various sessions, and it is their competent management and often stimulating interventions which played an important part in the success of the original symposium. In editing the papers for publication we also wish to acknowledge the contribution to our thinking made by John Bousfield, John Clammer, and James Danandjaja, and, specifically with respect to the Introduction, by Barry Hooker and John Jervis. The final responsibility for the Introduction is Ellen's, though its underlying revisionist positivism has been well tempered in places by Watson's postmodernism. We believe this editorial tension to have been productive. C. W. Watson Roy Ellen Eliot College, University ofKent at Canterbury

1

Introduction ROY E L L E N

Witchcraft and Sorcery as an Object of Academic Study in Southeast Asia IN ACCOUNTS of personal misfortune, beliefs concerning the imputed manipulation of mystical agency by other persons are as much a part of the contemporary Southeast Asian scene as the traditional. Malign magic and sorcery are reported in the literature (ethnographic, historical, autobiographical, and fictional) and are part of the routine experience of villagers and urban dwellers alike. From time to time stories appear in the press that bear witness to a persisting public concern with such matters (Lieban 1967:25; Slaats and Portier 1989).1 Yet, with the exception of a few monographs, such as Lieban's Cebuano Sorcery, the subject can hardly be said to be prominent in anthropological or other scholarly work on the region. And apart from Lieban's remarkable 111 cases and, in a rather different way, Golomb's work (1985) on multiethnic curing in Thailand, there is little case material— at least little which is published or accessible. Instead, accounts of witchcraft or sorcery are rather general, with few concrete illustrations of real-life accusations or confessions, or the social dramas that surround them. Barton (1969:65), for example, reports three Ifugao cases, but all of these are based on hearsay of happenings that occurred many years ago rather than events witnessed in the field. One is almost tempted to agree with Mary Douglas, when in desperation she speaks of witchcraft as a "private obsession of Africanists" (Douglas 1970b:xi). But why should the Southeast Asian literature not contain a body of coherent analyses comparable, say, with those available for sub-

1

2

ROY

CHINA

L

Lisu

S

Shan

T

Thai

(Durrenberger) (Tannenbaum) (Golomb)

NS Negeri Sembilan G BURMA

,—r r-

K

TAIWAN

ELLEN

Gayo

(Peletz)

(Bowen)

Karo Batak (Slaats

and

Portier)

LAOS"

M

Minangkabau

Kt

Kerinci

B

Banten

(Walson)

(Walson) (Nitibaskara,

Watson)

THAILAND

Luzon

N

Nage

(Förth)

Nu Nuaulu South

ACAMBODIAJ

China

°

(Ellen)

PHILIPPINES

Sea Pacific Ocean

MALAYSIA

0

BRUNEI Celebes

Sea Moluccas

Borneo

Sulawesi

o^ Sumbawa

tombok

Flores

r^v)

Seram

Irian J a y a

cs>

Oy

Southeast Asia, showing locations of case studies

Saharan Africa (e.g., Crawford 1967; Marwick 1965; Middleton and Winter 1963) or, more recently, for Melanesia (e.g., Stephen 1987; Zelenietz 1981) and early modern Europe (e.g., Favret-Saada 1980; Lamer 1981; Macfarlane 1970a)? Part of the reason for this must rest with the character of influential formative work on the subject. One thinks, for example, of Kluckhohn and Whiting's Amerindian models in the U.S. tradition, and the totemic presence of Evans-Pritchard for those trained in the British school. Certainly, the latter's Witchcraft, Oracles and Magic among the Azande provided a powerful exemplar to which other Africanists could turn. Given this, can we therefore conclude that had Evans-Pritchard undertaken his researches on the same subject in Malaysia or Thailand things might have turned out differently? Surely, it is too facile to say that the situation reflects the way in which European and American

Introduction

3

scholars have determined what the significant research problems are or the places where influential minds have chosen to do their fieldwork. We can perhaps draw a parallel example from the study of kinship. It is well known that Africanists have long emphasized descent in their models of lineage theory, whereas Asianists have emphasized alliance (Fox 1967:24; Schneider 1965). It would be absurd to explain this solely in terms of the predispositions of two groups of scholars, though it must certainly be partly true (see, e.g., Barnes 1962). But the emphases in the respective literatures go beyond what might reasonably be construed as constructions put on the text by interpreters, to what many would regard as objective features of the kinds of societies examined: the presence or absence of descent groups of a particular kind and of different material exchanges structuring marriage. It is of some significance, therefore, that Africanist studies of witchcraft and sorcery should have been more successfully applied (even though to some extent detrimentally) to Melanesian data, where sociological similarities might predispose us to expect them (Stephen 1987:1-3). The lack of prominence of witchcraft and sorcery in the ethnographic literature on Southeast Asia, and the overall low level of accusations compared with Africa or even Melanesia, also reflects in part its relative absence as a significant social problem, its characteristic presentation as a private rather than a public delict, and that in turn reflects the existence of alternative ways of dealing with misfortune and conflict.2 All of this finds corroboration in the relative absence of anti-witchcraft-type movements in Southeast Asia.3 The treatment of witchcraft and sorcery beließ is another matter altogether. In analyses of African thought, metaphysics, and rationality, witchcraft has become pivotal (Evans-Pritchard 1937; Gluckman 1944; Horton 1967), though it is ironically true that it inspired a generation of studies emphasizing the sociology of accusation at the expense of the structure of belief (Crick 1976:109-227; Douglas 1970c:xiv). In discussions of the cultural construction of thought in Southeast Asia the emphasis has inclined more toward other concepts held to be central in systems of belief and causality, such as Malay semmgat, vital mystical force (Endicott 1970; Skeat 1900), or adat, in one of its senses, "sacred"; on some framework held to organize cosmological ideas, such as symbolic dualism; or ideas derived from the locally dominant major traditions, such as Buddhist karma. Where we do find mentions of witchcraft and sorcery—at least in the older writing—it is as a footnote to illustrate the backwardness and primitive character, and indeed the "savagery," of indigenous thought. Thus, for W. W. Skeat, much

4

ROY E L L E N

influenced by Tylor and A. C. Haddon (and almost certainly, though indirectly, Frazer), though semavgat was not to be explained in terms of "gibberish theory," it was not inappropriately described as " a species of thumbling" (Gullick 1988:131,140-142, n80, n93, n95; HoodSalleh 1984:viii). A preoccupation with such matters emerges, too, in Dutch and English colonial novels and popular literature. It would seem that the situation described in the previous paragraph is also related to—on the face of it—a curious absence of reported sorcery accusations in the known colonial administrative records in Southeast Asia, and a lack of concern for its possible legal implications. Anthropologists working in Africa have often been able to draw upon extensive court and other government records, a consequence of witchcraft accusation and the employment of magic to harm others being given official recognition by a colonial administration as a penal offense (e.g., Schapera 1970:108-109). In some places the legal construction of the offense outside a colonial context is reflected in the existence of extensive indigenous trial records (ibid., p. 108), which has sometimes enabled extensive scrutiny of witchcraft cases brought before local courts (Reynolds 1963). In Northern Rhodesia it way a criminal offense to impute witchcraft or sorcery (Marwick 1965:18), and Marwick (1965:205 and passim) informs us that Cewa chiefly courts were thoroughly involved in litigation in this connection. This was not always the case, as when the establishment of a British Protectorate in 1900 led to Barotse courts losing their power to try cases relating to witchcraft, "except with special permission" (Gluckman 1967:60). The African experience was repeated in the New Guinea Native Regulations as early as 1935, which, while legally defining sorcery as "deceit," accepted that it was deceit with unacceptable social consequences and therefore had to be punished (Zelenietz 1981:12). But in Southeast Asian laws and court records for the colonial period it barely seems to feature. The virtual absence of any reference to witchcraft or sorcery in Southeast Asia in this connection must in part have been determined by the way in which colonial laws with respect to native subjects were formulated: formal Islamic law (Shariah) in Malaya, and the complex edifice of adatrecht in Indonesia, stressing the ideal and eternal at the expense of the actual and transient (Ellen 1976:317-319; Ellen 1983:50-54). In Malaya, the British were obliged to refrain from political interference in religious affairs and anything that might be construed as meddling (e.g., scholarly research). Reliance on Shariah meant that sorcery accusations seldom came before the courts. In the Dutch East Indies, by comparison, though the application of adat law was less

Introduction

5

constrained by perceptions of Islam than by modern Western values, this too may go some way toward explaining the exclusion of references to sorcery. Thus, in the British- and Dutch-controlled territories of Southeast Asia, laws against sorcery were never seen, as was the case in New Guinea (Zelenietz 1981:12), as a progressive tool for social engineering. Although morally reprehensible, malign magic was legally invisible. Whereas in Africa and Melanesia colonial governments recognized its social importance by making it illegal while at the same time claiming it to be illusory, the British and Dutch in Southeast Asia denied its existence in law while consequentially understating its potential for social disruption.4 With independence, the legal codes of the previous colonial governments continued to operate. Increasingly, however, postcolonial governments and their subjects found that they could not accept many of the underlying jurisprudential values. This is well illustrated with respect to sorcery in Indonesia. The Dutch colonial government had early attempted to foster an enlightened modernism in such matters, as reflected in the rationalist critique of superstition adopted by the Balai Pustaka discussed here by Watson. They had, one supposes, anticipated that in the fullness of time such education would lead to the demise of magic. A similar approach appears to have been evident in the attitudes of postindependence Indonesian administrations, though in recent years the paradigm appears to be showing signs of dramatic leakage, almost to the point of moral panic. If ordinary people, as well as politicians and government officials believe in the efficacy of malign magic and hold its imputed practitioners to be wicked, then a legal system that fails to deal with a pressing social problem is woefully inadequate. As Slaats and Portier emphasize, there is a growing readiness among Indonesian lawyers to hear complaints, and there have been attempts to change the law so as to make the practice of magic an offense. The pursuit of an effective legal framework for dealing with mystical aggression continues in Indonesia, as indeed it does elsewhere (e.g., Zelenietz 1981:12). Understandably, in the context of modern European legal traditions this raises serious problems in terms of the acceptability of evidence, as when, for example, claims are made (as documented here by Nitibaskara) that it is widely assumed that certain magical practitioners make false claims for their powers or misuse them to harm others. And when, as in Indonesia, purported instances of malign magic become media events, pursued and inflated by scurrilous hacks and a prurient readership, we might reasonably ask—as did Mark

6

ROY E L L E N

Hobart at the symposium on which this book is based—whether having entered a wider public domain, extracted from their immediate context and proximate causes, the phenomena have not been fundamentally transformed?

A Question of Terminology It is well known that the supernatural harming of others is sometimes considered to be the consequence of an inherent faculty or a disordered personality, and sometimes the learned and conscious manipulation of objects, spirits, and words. This is the convenient and conventional distinction between witchcraft and sorcery, confirmed as such through the definitions of Evans-Pritchard. It is equally well known that in many cultures the distinction is hard to maintain and often thoroughly misleading (Marwick 1967:232-234; Turner 1964:318-322). In some parts of Southeast Asia witches and sorcerers are conceptualized as separate entities and terminologically distinguished (e.g., Lisu). Whereas witchhood constitutes an involuntary and passive property of the person, a sorcerer is an active, voluntary participant in an occupational role. Something like this distinction appears to exist for the Nage on Flores (Forth 1989) who believe that human beings may be metamorphised into witches. But more often than not there are no terms to distinguish the two, even where (as among the Cebuano) the notions are more or less conceptually distinct (Lieban 1967:65). Moreover, it is frequently the case that there is no generic term glossing witchcraft and sorcery collectively or sorcery as a whole. Rather, we are inclined to find a number of terms (sometimes a large number) for particular subtypes (e.g., Barton 1969:63; Lieban 1967:48-75). Occasionally, the term for a specific subtype is extended and used as a generic (as in Cebuano bamn0). It is common for the conventional diagnostic features to be radically recombined. Thus, in the Nuaulu case, while the capacity to harm others mystically requires an inherited physical trait, it must be activated through the manipulation of words and objects, and through consultation with a cooperative ancestral spirit. Balinese ¿^«¿—generally described as "witches" in the literature—become so by studying the writings on the subject if they have first acquired a passive knowledge through intense asceticism, and then may only practice with the express permission of the supreme goddess Durga (Howe 1984:215). Moreover, sorcery is not universally "black," in the sense that many accept that harm may be in other respects legitimately caused to pre-

Introduction

7

vent a greater harm, or in the sense that the intentions of a witch or of magical manipulation may be entirely benign. Witches may variously be the epitome of evil (as in the case of the Balinese Rangda), may be inferred as being generally unpleasant from beliefs (such as the Nage notion that being transformed into one constitutes a punishment [Forth 1989]), or may be considered irritatingly playful but relatively harmless (Cebuano). Golomb (1985:96) has described at length the overlap between love magic and sorcery in Thailand: the former harmful in the sense that it forces someone to act amorously who otherwise might not wish to, while the latter may well be undertaken to right a wrong. Practitioners often occupy otherwise socially acceptable and morally neutral all-purpose roles (Malay bomoh, Thai moo, Javanese dukun), where any practitioner may have mastered a number of specialities—bone-setting, midwifery, massage, healing, the purveying of spells, divination, exorcism, herbalism, sorcery, and so on. In no sense can these general ritual experts be construed as "gangsters of the supernatural," to use Stephen's (1987:7) evocative (but one suspects ultimately misleading) description of African sorcery.5 The general point of all this, of course, is that any assessment of a purported action will wholly depend on the position of an observer in relation to an accusation. It is this which determines who the victim is, what the intentions of the person accused were, and therefore what thoughts and practices are deemed malign. Events will be constructed and interpreted differently depending on whether the active self is practitioner, victim, accused, or quite unrelated to what is going on. There is no particular reason to believe that in most of these respects Southeast Asia is significantly different from any other part of the world. Where we think it is different—though maybe only in degree—is in terms of the construction of the sorcerer as a facet of the more general role of curer-magician and in the existence of sorcery as a part-time vocation. In other words, some practitioners of this specialized knowledge may make a modest living out of seeking clients and offering their services for hire, though they may not think they are doing so with malign intent. This is not, of course, true everywhere. We know that in some societies sorcerers only exist in the minds of those who accuse them—that is, no one would ever willfully engage in the craft, or if they did so, would ever admit it. Thus, Resner and Hartog (1970:375) found no Malay or outgroup person who would admit to performing sorcery. Among the Siamese Thais of central and southern Thailand, tales abound, knowledge is widespread, but locating actual malevolent sorcerers in quite another matter. But in some Southeast Asian societies

8

ROY ELLEN

persons are prepared to identify with the role of sorcerer, even if they will seldom—if ever—admit to particular acts. Where this is the case, clients and interested observers may even form judgments as to the authenticity or phoneyness of a practitioner (Nitibaskara, Chap. 7). Southeast Asia is different from Africa in at least one other respect (though this is also something it shares with Oceania), namely, in that what we conventionally call sorcery is infinitely more common than witchcraft, to the extent that sorcery has come to serve as a dominant paradigm in explaining the use of supernatural agency by persons to harm other persons. Witchcraft—in its narrow sense—is the exception. For this reason, we here use the term "sorcery" to include witchcraft, and only employ the word "witchcraft" in its specific sense as and when it is necessary to make the time-honored conceptual distinction. This, of course, is the opposite practice to that adopted by the majority of Africanists, and those generalizing on the basis of (mostly) African material.6 But, as with the Africanist consensus, it is only a convention and does not have the force of an analytical category—as we shall see.

Etiology, Cosmology, and Epistemology Following Jung and others, it has been plausibly suggested that the primal character of witch and sorcerer images can only be explained if we see them as spontaneous creations of the imagination latent in all cultures. The differences across the cultural continuum are therefore no more than variations in the way in which the unconscious is translated and transformed into specific social behavior (Stephen 1987:297). Both the images themselves and the deployment of images to make sense of and attempt to alter experience are as much a part of the maintenance of cosmic order as any other aspect of what we call religious belief. How the problematic of sorcery and witchcraft is thereafter constructed depends on whether the subject is approached from the angle of the practitioner, victim (client), accuser, or some other third party, and what this might tell us about the representation of power. Having said as much, it is now necessary to examine whether the categories of "malign magic," cognate notions of rationality and causality, associated representations, and terminology are differently constructed in Southeast Asia compared to other parts of the world, and, if so, to see in what ways this might help to explain the relative inattention ethnographers have paid to the subject. All witchcraft and sorcery beliefs seem to be underpinned by some form of folk-essentialism: the idea that actual attacks represent the con-

Introduction

9

sequence of the existence of some general cosmic power which may be manifested in various different ways, only a few of which can be glossed by terms such as "sorcery" or "witchcraft." Thus, the primal mystical concept in Malay thought, around which all else is imputed to revolve, has always been represented as senuwgat. We have already suggested that the ubiquitousness of such concepts may well have been one of the reasons why sorcery and witchcraft as an Africanist or Oceanist might perceive it has been peripheralized in Southeast Asia. But it might be argued that the mode of engagement between the Orientalist and native Other does not fully explain what is going on. There are grounds for believing that in many parts of Southeast Asia the symbolic construction of the cosmos, the social world, and agency support this view. The notion that sorcery is no more than an aspect of some unitary cosmic force appears to have had two seemingly paradoxical consequences: first, to make it less obtrusive as a separate phenomenon, and second, to bring it right to the core of the cultural expression of religious ideas. Thus, Thai shadow-play performers are often suspected of possessing powerful magical skills—indeed, chaatrii means sorcerer; the main figure in Kelantanese Malay theater is a bomoh, or specialist in curing magic; and in Bali the Rangda dance-drama, familiar from its representations in the tourist trade, is indigenously salient as well. But while such ideas may theoretically unify sorcery beliefs within some greater whole, in the lives of most ordinary people the immediate concern is with agency, how it is that such power—and sorcery is, willynilly, entwined in the exercise of power—can be channeled for malign purposes. We may distinguish four kinds of agent. 1. The power may be innate in certain individuals, which is what we usually mean by witchcraft. Such power may be exercised privately or be apparent in a glance or gaze—the "evil eye." In some places you may become a witch by killing a witch (as in Lisu), and it is quite widely believed that the power may be inherited. Where the latter is the case, and where descent arrangements permit, it is not surprising that we occasionally come across witch clans and lineages as described for the Lisu by Paul Durrenberger (Chap. 3; but see also Davis 1984:76). 2. Alternatively, the agent may be an independentfree-rangingspirit that is harnessed by a sorcerer for the purpose. There appears to be a wide consensus within Southeast Asia as to which spirits are particularly suitable, often those of a stillborn child or of a person who has died a violent death. The borderline between these first two kinds of agent may not always be clear, and some peoples (e.g., the Nage) treat witches as a special kind of spirit.

10

ROY ELLEN

3. Third, the power may be harnessed through the direct manipulation of objects, characteristically personal leavings, images of the intended victim, and animals. Where animals are involved, these are usually insects—e.g., Cebuano barang, which are regarded as being able to enter the body (Lieban 1967:1). This is similar to Malay bajang, conjured up from the newly dug grave of a stillborn child, fed in a stoppered container, and released when required. These may take the form of particular animals, including insects (Winstedt 1961:25). Alternatively, as in Ifugao ayak (soul-stealing), it is the soul of the victim which is metamorphosed into an insect, in order that it may be more easily trapped to drink rice wine and imprisoned in a bamboo container to bring on death. Evidence of overlap between elements (1) and (3) is common, as in the Balinese belief that many witches (for example, the leyak) may actually be animals, while many witches and sorcerers are thought to turn into animals to perform their work. It is not surprising that sorcery should be seen to act through animals—and particularly through insects and birds—as in the absence of a germ theory they provide a material explanation of infection across space and time, a means of physically connecting sorcerer with victim. Where there is no such physical trace, the active agent of sorcery is often represented as wind. In Thai, this is, appropriately, also the most common word for illness (Golomb 1985:138). Among the Cebuano (Lieban 1967:148) germs and parasites are sometimes conceptualized as the potential disease-inflicting instruments of sorcerers. 4. Fourth, the power may come through utterance of certain words and spells. Attack through utterance alone (especially when it is public) is, of course, a curse; but in most cases it is believed that effective sorcery entails some combination of these various elements. The precise combination is extremely variable, and it is this which hampers the discovery of convenient cross-cultural typologies. There is variation in the extent to which sorcery is regarded as evil. Everywhere it is associated with bad death, but a bad death is not necessarily a consequence of evildoing. Any death of a Nuaulu child is a bad death, but its cause may be reckoned as ancestral retribution for failure to attend to some ritual particular. Ancestors who act in this way are not evil, they are simply acting as one would expect them to act in the circumstances: ensuring that proper cosmological order is maintained by exerting sanctions on the living. Similarly, the Nuaulu sorcerer is not intrinsically evil, just powerful; and the morality or legiti-

Introduction

11

macy of a putative act of sorcery must be evaluated according to its context; it cannot be legislated for in advance. After the same fashion, some Cebuano regard witchcraft and sorcery simply as "different, not natural"; and Cebuano witches such as buyagan (as opposed to sorcerers—and despite Christian influence—are thought to be relatively benign. Often, they are considered to be not deliberately malicious, to cause only minor ailments such as skin complaints, and to be as much victims as those whom they purportedly attack. For many Thai, sorcery accusations are a matter for embarrassment rather than anger. Elsewhere sorcery is definitely evil. What all o f the religious traditions have in common is the conceptual opposition o f sorcery to all that is normal and good. On Bali, witchcraft is associated with the goddess Durga (the demonic form o f Uma, who is in turn the consort of Siwa), the personification o f evil and Bali's chief léyak. Witches are opposed to the way o f the gods and righteous people, and are chronically disposed to be greedy, jealous, angry, malevolent, and thoroughly obnoxious. Moreover, structurally, they are the opposite o f all that is normal: they are associated with the left rather than right, night rather than the day, animal rather than human; they dance upside down on graves at midnight and enjoy everything ordinary Balinese find revolting (Howe 1984:212-217). In such cases, heavy punishment o f a sorcerer for alleged misdeeds is often felt necessary, as among the Batak (Slaats and Portier 1989), and we might well expect pressure on the state to provide a legal mechanism to satisfy moral outrage.

Malign Magic and the Major Religious Traditions I f it were ever so, it is now no longer tenable to claim an intrinsic connection between witchcraft or sorcery and animism, or even "primitive" thought. Abundant comparative ethnographic and historical data attest to the importance o f such beliefs in Christian, Muslim, Buddhist, and Hindu contexts. Moreover, some o f the finest studies of the subject are of societies much influenced by, in particular, Christian beliefs: from Africa (Crawford 1967; Marwick 1965), Europe (Baroja 1964; Favret-Saada 1980; Larner 1981; Macfarlane 1970b), and from Southeast Asia (Lieban 1967). It is, however, necessary to ask whether belief in canonical religions has any general effect on the expression o f witchcraft or sorcery beliefs, and whether assimilation into particular religions has any particular consequences. It is now absolutely clear that social change itself may prompt an increase in accusations, and that religious change is an indivisible part o f

12

ROY E L L E N

this. But in terms of the dynamics of interacting belief systems, an encounter with any religion with exclusive claims and moral authority would seem either to force witchcraft beliefs into the inferior and adversarial complement of a dualistic moral order or to encourage further such dualistic notions as might already exist. Thus the spirits with which Cebuano sorcerers engage are characterized as followers of Satan, and sorcery as the work of the Devil (Lieban 1967:21). Healers treat sorcery as a confrontation between G o d and the Devil where historically Christian missions have defined spirits of the pagan religious world as representations of the Devil. The Cebuano invokes ingkanto for sorcery, God or the Saints for healing (p. 40). Indeed, the Cebuano ethnography suggests a movement from beliefs where sorcerers manipulate various free-ranging spiritual forces to one where they are regarded as entering into a compact with the personification of evil (cf. Larner 1984:80). Such a transition is obviously aided where there are preexisting beliefs in a compact with individual spirits (Cebuano, Nuaulu), and it is tempting to conclude that wherever institutionalized religion has to cope with the anarchy presented by sorcery, it does so by seeing it as an aspect of the action of some general personalized singular evil force, or gods in pantheistic religions. Thus, Durga enters into a compact with certain Balinese sorcerers, as Diana and Hecate are reputed to have done in classical times (Baroja 1964:39-450). Unfortunately, the picture is not always so clear-cut. Thus, some Cebuano sorcerers will sorcerize for others, saying that "life is precious" or that " i t is God's c o m m a n d " (ibid., p. 28). Cebuano healers are regarded as having access to God's power (ibid., pp. 32-33), the implication being that G o d allows sorcery to work. Indeed, viewed globally, Christian theology seems simultaneously to deny the ontological possibility of malign magic while attempting to control (forbid) it in the real world. The position is not dissimilar among Muslims. Thus, some Thai Muslims (especially reformists) follow the Koran in saying that all spirits are a manifestation of Satan, while others say that this is only true for some, and still others claim that even spirits performing errands for sorcerers can be champions o f Muslim piety by, for example, refusing to possess haram Buddhists (Golomb 1985:217). Golomb reports that members of opposing Muslim factions may occasionally employ (usually a Thai) Buddhist sorcerer against each other (ibid., p. 16). Despite the forbidding of such practices by some religious authorities, it is an intrinsic part of the constitution of Islam in Southeast Asian societies. For, as Bowen demonstrates, the Gayo embed magical beliefs in a dis-

Introduction

13

tinctively Islamic cosmology and etiology, and are happy to construe counter-sorcery as an acceptable morally neutral return. The Tauhidic worldview does not attribute evil to a single source, and therefore, in contrast to Christianity, would not appear to have the same polarizing effect. Similar constructions could be given for several other societies described in this volume. Even so, it would seem to be more difficult for Muslims to reconcile animism with religious observance than Buddhists, for whom supernaturalism is almost a morally neutral matter (Golomb 1985:106). Like Muslims, Buddhists can point to relevant scriptural passages for final authority on the existence of spirits. In Burma, monks are reported as being prohibited from dabbling (Spiro 1970:160), though in Thailand at least they may serve as exorcists, seeking monasteries in regions where the moral balance is in need of redress (Golomb 1985:213). What makes such beliefs in sorcery ultimately feasible in such systems is the relativity of evil, the fact that what is malevolent behavior for one person is not for another.

The Social Epidemiology of Sorcery In looking at the significance of sorcery beliefs in Southeast Asia it is necessary to distinguish between shared beliefs and observations of specific events. In some places there is little evidence for either. For example, the Kalinga of northern Luzon are reported to have little belief in the efficacy of sorcery and find it difficult to take it seriously (Barton 1949). Monographs and other accounts of disputes and their settlement, of magic and metaphysics, which we might normally expect to contain references to sorcery, either contain nothing (e.g., Schlegel 1970) or very little (Davis 1984). But as we have already seen, absence of reference may have as much to do with scholarly preoccupations as with anything else. Elsewhere in the region sorcery is spoken of as a potent cosmic force, but accusations seem to be rare; in other words, there is a disjunction between belief and practice. This is the case, for example, among the Nuaulu (Ellen, Chap. 5) and in Bali (Howe 1984:213). If beliefs about the role of malign magic are common, and part of the explanatory repertoire of most Southeast Asian peoples, why should specific accusations be so rare? In places both beliefs and instances of specific accusation are widespread, as in Negeri Sembilan, Patani, and in the Lao-speaking areas of northern Thailand (Peletz, Chap. 9; Golomb 1985:114-116). There is, then—although we may be skeptical of placing too much

14

ROY E L L E N

reliance on the omission of data or on speculative attempts to quantify its significance—some reason to believe that sorcery beliefs themselves and the level of accusations are patchily distributed in Southeast Asia. In this case, what regularities can be discerned in that distribution and how might they be accounted for? Is it likely that the variation can be explained everywhere in a similar way? Two things are certain: that we need to make a distinction between explanations of the incidence of accusations and the variable character and intensity of sorcery beliefs, and that variation in significance may have less to do with the cultural construction of the subject in particular populations than with critical internal cleavages, the circumstances of individuals within a locality, or social patterns cutting across ethnic lines. Before even beginning an attempt to locate social factors that might explain degrees of incidence, we must establish a set of locally appropriate ethnoepistemological parameters. Both levels of accusation and the cosmological prominence of sorcery beliefs greatly depend on the availability of other means of explaining misfortune: the extent to which nonmystical reasons are acceptable, the perceived function of ancestors, of a supreme deity, lesser deities, various spirits, and other cosmic forces. Thus, I have explained the low levels of Nuaulu accusation in terms of the structural priority accorded to ancestors in coping with all misfortune. Elsewhere, as among Malaysian Peranakan (Clammer, personal communication), disinterest in sorcery stems from a systemic view of evil in which it is not necessary to attribute misfortune to the ill will of individual persons, as evil dwells in the very interstices of the cosmos; and where such alternative explanatory paradigms are dominant, sorcery (as on Bali) becomes a residual category. Our ignorance of how sorcery functions in much of Southeast Asia would make premature the setting up of typologies that correlate genus of malign influence with the social profile of those accused and the structure of the social formation in which they operate. Indeed, given what we do know, the prospect of ever being able to do so may seem rather remote. Unlike precolonial Melanesia or much of sub-Saharan Africa, Southeast Asia has long been typified by extreme ethnic interpenetration, complex patterns of cultural contact, and intricate political configurations. The kinds of complications that arise are well exemplified by brief reference to the classic insider/outsider contrast. Much of the competent literature on Southeast Asia stresses the extent to which outgroups are accused of sorcery: Buddhist Thais accusing Karens (Keyes 1980:6, nl8), Kelantanese Malay accusing Thais (Gimlette and Thomson 1971 [1939]; Resner and Hartog 1970),

Introduction

15

ethnic Thais accusing local Malays and Thai converts to Islam (Golomb 1985:270), Malays accusing local Chinese (Colson 1971:31; Provencher 1975:142), and so on. But such general statements can be misleading on several counts. First, we should not confuse general assertions with specific accusations. General assertions of this kind are, of course, interesting in themselves, in that they tell us something about how different groups o f outsiders are perceived, but they do not necessarily reflect lines of social tension. Nuaulu on Seram will, in general terms, speak of various outgroups as having reputations as sorcerers, but members of such groups are rarely encountered and even less often accused of being responsible for a specific misfortune. Even local outsiders (Muslims, Christians, Ambonese, Butonese) are not among those most specifically accused, who tend to be insiders. Another complication with the ethnic outsider explanation—especially in a comparative dimension—exists in those ethnically plural locations where sorcerers are available for hire. Here, although ethnic outsiders may have been the agents through which a misfortune resulted, these agents are under the instructions of others—often ethnic insiders. Thus we must distinguish views about those who have a reputation for sorcery and those who are most likely to use sorcery or sorcerers to harm us. Although the most powerful magic is possessed by those living in the social margins (e.g., Khymers), in Thailand sorcery is directed predominantly at members of one's own ethnic group (Golomb 1985:201, 209). In those places where sorcery is regarded as the speciality of particular (marginal) groups, the extent of specific accusations may also be affected by the availability of suitable outgroup practitioners (Golomb 1985:211; Textor 1960). When accusations are directed inwardly, those accused can be interpreted as reflecting obvious structural tensions within the community: between chief and commoner, contestants in love, cowives or wives and concubines (Golomb 1985:240); or as reflecting views as to those who might have resorted to sorcery both because of their power (Nuaulu) or because of their powerlessness, such as the old Kalinga women described by Barton (1949).

Sorcery and Social Change Patterns of accusation and the construction of the subject change over time. In places where sorcery was once confined to chiefs or other authority figures it is now available to all; from being a means of controlling the status quo it may become a means of upsetting it. On the

16

ROY ELLEN

other hand, when political centralization and institutional religion emerge where they have previously been absent, sorcery as an informal means of egalitarian social control may become a threat to the authority and livelihood of a priesthood and leadership dependent on orthodoxy, control of belief, and of other specialists in supernatural matters. The existence of sorcerers represents a challenge to that power, as they are autonomous and unpredictable experts in the supernatural. Changes we see in the constitution of sorcery beliefs in such societies invariably reflect these contrasting scenarios, particularly in both cases the loss of the indigenous legitimacy of sorcery practices. It is also possible that as the role of sorcerer becomes delegitimized, so the roles of witch and sorcerer merge. Moreover, as has been long recognized, the uncertainties of change may themselves stimulate accusations, and for Southeast Asia there is some evidence that these tend to rise where other "traditional" forms of social control have broken down and not yet been replaced by alternatives, usually agencies of the state (Lieban 1967:124), where individual redress and violence are not (or no longer) socially legitimate (Peletz, Chap. 9), and where traditional political authority encompassing control of mystical power has passed to a more individualistic regime where magicians retain social power for nosological reasons. This position is exemplified by a comparison of Lombok and Java by Sven Cedderoth presented at the symposium but published elsewhere (Cedderoth 1990). For the modern period, another factor thought to have a bearing on the incidence of accusation is degree of urbanization. The standard assumption of the partially informed used to be that urbanization was part of that process crudely designated modernization which relendessly undermined primitive beliefs, and of course it is the case that in many areas sorcery is more important in the lives of rural people than in those of the urban masses. But sorcery is in no sense restricted to rural settings. It is clearly alive and well among city-dwelling Indonesians and in urban Thailand, where Golomb suggests its high profile reflects an increasingly competitive social environment. Notwithstanding this, however, Lieban (1967:137), in one of the few detailed empirical tests of whether urban living results in an increase or decrease of accusations, concludes that in Cebu City at least accusations are less prevalent pro rata than in the rural areas. He plausibly argues that high population concentrations may make sorcery more visible to the ethnographer, but we should not be misled into thinking that they have increased as a result of living cheek-by-jowl, greater relative deprivation, or "an increasingly competitive environment." The reason why

Introduction

17

accusations decrease in urban areas, argues Lieban (1967:149), is because there is greater poverty—and specifically less landed property to argue over—and fewer obligations to cooperate in economic matters. The conditions for increasing accusations are met rather in those rural populations where there remains a strong residue of traditional expectations but also some erosion of traditional sanctions. If this is so, then it goes some way to explain the high incidence in those parts of the lowland Philippines that underwent early structural dislocation and religious conversion under Spanish colonialism, but (partly as a consequence of poverty) rely heavily on traditional modes of therapy and coping strategies. The Lieban hypothesis accounts for the decline of incidence of accusation in contemporary urban settings, but it does not settle the issue of the persistence of belief, why it should be that sophisticated opinion formers in developing societies should continue to believe in the mystical transmission of ill will (e.g., Golomb 1985:205). Our own view here is that anthropologists have never succeeded inframingthe problematic properly, happy to reinforce their own rational, Western, and agnostic prejudices with old Frazerian stereotypes. For if we can accept that rational thought, modernity, and religious belief are compatible, then why not witchcraft or sorcery? That witchcraft in northwestern Europe virtually died out when it did has misled many into believing that it was an inevitable consequence of that new mode of thought which accompanied the rise of capitalism and modern science. The existence and revival of "neopaganism" and witchcraft among computer programmers in Tufnell Park would seem to cast some doubt upon this long-cherished assumption (Luhrmann 1989).

Sorcery in Southeast Asia and Anthropological Theory It is two decades since the last major theoretical stocktaking concerning witchcraft and sorcery. Indeed, it is almost as if theorizing had been effectively abandoned in the early seventies with the publication of three definitive volumes, all essentially Africanist in orientation (Douglas 1970a; Mair 1969; Marwick 1970). This impression of abandonment may owe something to a feeling of sheer exhaustion after a decade of active analysis and debate, but we suspect also a degree of corporate professional satisfaction that all the major questions had been addressed and most of them resolved, at least in the light of what was then known. But does the apparent consensus of the early seventies represent no more than complacency, and is it the case that the anthro-

18

ROY E L L E N

pology of sorcery has now to be substantially (if not entirely) recast? This introductory essay concludes, therefore, by examining the extent to which the Southeast Asian literature might motivate us to reassess the underlying theses in this body of work. There are at least some grounds for believing that, for reasons embedded in both the history of anthropology and in the ethnography of the region, studies of sorcery in Southeast Asia might allow for the rapprochement of the various approaches in a more sensitive way, and go beyond simple structural and typological explanations. Theories pertaining to sorcery are generally reckoned to be of three kinds: intellectualist (sometimes called cognitive) theories; sociological theories, which explain the discontinuity and incidence of accusation (these are mainly social control theories); and affective theories. Intellectualist theories are those whose interest in sorcery is as a window into certain kinds of worldview, ways o f explaining, "systems" of rationality. N o t content to ask questions about incidence, they seek to go beyond this and ask why people should hold such (implicitly " a b s u r d " ) beliefs in the first place. Until the 1960s, sorcery in Southeast Asia was represented by anthropologists in terms of theories of this kind, though owing more to Tylor, Frazer, and Lévy-Bruhl than to Evans-Pritchard. What we have found remarkable in our own survey of the literature is the extent to which this earlier paradigm (though perhaps dressed up in fashionably new ways) is still prevalent. One reason for this may be that Zande interrogative casuistry (why me? why here? why now?) is not particularly appropriate, as is evident in the remarks made here by Michael Peletz on Negri Sembilan. Sociological theories include such well-worn hypotheses as those which state that fear of accusation prevents people from behaving in antisocial ways likely to provoke the anger of a witch or sorcerer, or that people are persuaded not to act like witches lest they themselves be accused of being one. Such assertions (though they are intuitively appealing) cannot be effectively measured, and it is never entirely clear that when different ethnographers speak in general terms of the rate of accusation, they are always measuring the same thing. It is on account of this that Ellen, for example, argues for a disaggregation of accusations as between trivial and significant, on the ground that in cases where no real attempt is made to identify and pursue an offender structural explanations are less likely to be as plausible as intellectualist or affective ones. Measurement is also impaired where, as on Bali, people are labeled as witches without being accused of bringing about specific misfortunes, and where people are unwilling to name names when mis-

Introduction

19

fortunes attributable to witchcraft do occur, out of fear of drawing attention to themselves. But, granted such problems as these and the general circumspection required when dealing with functionalist explanations, there is a range of more specific hypotheses which can, in principle at least, be tested, given suitable quantified data. Certainly, we should not feel cowed by Turner's observation that witchcraft is a rather clumsy method of effecting social control, since where it is preemptive it can be engagingly subtle. Looking specifically at studies in the Southeast Asian field, such issues have only been effectively tackled in the work of Lieban and Golomb, both of whom—significantly—have worked in market-oriented peasant societies. By effectively we mean that only these researchers have as yet made any serious attempt to supply the kinds of data required to make any such hypotheses more than just plausible. From the work of Lieban, Golomb, and a few others, we can extract a number of hypotheses which suggest, variously, that accusations are most frequent where sanctions are in terms of minor punishments (and infrequent where the consequences are grave) (Golomb 1985:114116); where there is no convenient outgroup to scapegoat (Golomb 1985:114); where sorcery is a private rather than a public (criminal) delict; where no other means of explaining misfortune is structurally more appropriate (Peletz, chap. 9); and where there is unequal access to resources, as between town and country. While it is difficult to prove that any of these hypotheses are valid widely for Southeast Asian data, it is quite clear that some of them definitely are not. For example, the idea of the sorcerer as anti-image will not work where an individual's reputation is actually enhanced by the public knowledge that he is a sorcerer (Nuaulu). In general, the varieties of concrete circumstances in which misfortune is mystically explained suggest to us that no formulaic sociological hypotheses will ever go far in comparative study. Affective theories are those that see sorcery as a means by which physically ill or mentally distressed individuals cope with their predicament, or a culture-bound projection of "neurosis," "schizophrenia," guilt, or whatever. This approach is usually associated in the first instance with the work of Clyde Kluckhohn (1944) and Beatrice Whiting (1950), and latterly with Lévi-Strauss (1963), for whom sorcery is a verbal language for dealing with affliction. In other words, it is an aspect of that body of strategies which people employ to prevent and alleviate sickness. In this respect, the identity of perpetrators of misfortune is not always required knowledge, as Peletz indicates for Negri Sembilan and Lieban for the Cebuano (1967:75-76); and in such cases

20

ROY ELLEN

reform of the law to bring sorcery cases within its orbit is probably beside the point. In the context of Southeast Asian ethnography, this highlights a difference of emphasis that has major implications for theory. The Africanists, for all their interest in modes of rationality, have for the most part firmly placed the study of witchcraft in the sociology of social control. This was equally true of the early Melanesianists such as Fortune (1932), Malinowski (1926), and their successors, and is a ready indicator of the major preoccupations of the anthropology of their time. The postwar accounts of sorcery in Southeast Asia are linked to a different problematic and set of practical issues related to the diagnosis and curing of sickness. The scene had been partly set by the prewar work of people like J. D. Gimlette (Gimlette 1929; Gimlette and Thomson 1971 [1939]) and Dutch medical officers such as J. H. F. Kohlbrugge (1907). Colonial practice, contemporaneous anthropological fashion, and certain apparent features of Southeast Asian societies themselves did not provide fertile ground for a discourse on witchcraft rooted in social control. Instead, witchcraft and sorcery reemerged within the discourse of medical anthropology, which was, significantly, becoming an increasingly important subdiscipline and paradigm from the late sixties onward. Thus, Lieban's excellent case material allowed him to link up with medical diagnoses for victims in his sample, while he was also able to show how sensations of sorcery patients, as well as visible symptoms of their maladies, may correspond in highly specific ways to the expected results of certain sorcery procedures (Lieban 1967:109). In effect, he was able to draw together the sociological and medical paradigms to show that "the frequency with which Cebuanos perceive sorcery as the cause of illness or death is a function of their medical situation and the state of their social relationships" (Lieban 1967:5). He thus inadvertently takes his cue from Turner's (1964) suggestion that there might be a correlation between purported instances of accusation and morbidity. In their original and unreconstructed form, all of these theoretical strands have one major characteristic in common: they are literalist. That is, they start from the assumption that an accuser, victim, and his or her audience generally agree that describing someone as a witch or a sorcerer has the same truth value as saying that he is a plumber or a tax inspector. But what makes sociological analysis difficult is the fact that witchcraft and sorcery are also widespread sources of metaphors for talking about other things. The Nuaulu, for example, may use sorcery

Introduction

21

as a means of coping with quite trivial hurts and misdemeanours without ever intending to follow through to an accusation; a particular choice of words or idiom simply defines and gives moral significance to an otherwise inexplicable happenstance. The terms are widely part of a figurative discourse that includes insult and derogation, and that may have no necessary mystical implication. Moreover, when people such as the Nage use "witch" as hyperbole, it sometimes makes it difficult to discriminate "real" accusations from metaphorical ones. Witchcraft and sorcery in all societies are potent sources of expressive language, symbolism, and allegory, which (as Watson points out) may be as important as actual accusations and cannot be reduced to specific sociological or intellectualist theories. And where this happens it can hardly be said that witchcraft and sorcery constitute a separate discourse at all; rather, they provide a repertoire of figures of speech whose significance and interpretation alter depending on a particular substantive discourse or domain. Thus, where we are dealing with specific accusations with clear histories and consequences we would be well advised to situate analysis not in some abstract realm of "belief but instead in the workaday language of, say, land tenure, prestige conflict, commercial rivalry, or sickness. This is not to concede the arguments of utilitarian reductionism, nor is it an admission that witchcraft and sorcery are simply ways of talking about other things. It is rather a recognition that the phenomena must always present themselves in a particular context and that, in each of these, literalist and figurative uses of the idiom may seem equally appropriate. Indeed, some might even be prepared to acknowledge that they coexist simultaneously. The vital ethnographic question then becomes under what circumstances might the balance between the two be said to alter, and how might we interrogate the boundary between them when the reality for observer and observed alike must inevitably be construed through metaphor.

Notes and References 1. I say persisting. On 29 August 1903, the Philippine newspaper La Democracia reported the case of men having been hung for killing a supposed witch; and Jose Nunez in El Bencimiento for 9 December 1905, has an entire article on the subject (see Blair and Robertson 1973:43, 311-319). There is no reason to think that these accounts are atypical. 2. But see Nitibaskara and Slaats and Portier in this volume (chaps. 7 and 8). 3. In only a few contemporary Southeast Asian societies is sorcery regarded as a public offense. In the past the picture may have been different, at least in some places. Forth notes that among the Nage (Flores) witchcraft was a matter

22

ROY ELLEN

for public accusation before the imposition of colonial rule, and that organized witch killing here and in other parts of Nusa Tenggara Timur was associated with the social trauma accompanying Dutch takeover. Among the Ifugao the rehearsal of appropriate magic (regardless of its efficaciousness) was sufficient grounds for inflicting the death penalty, and even a reputation for sorcery could lead to death on the spot (Barton 1949:64-65). Perhaps, also, the general lack of anti-witchcraft movements may reflect the failure of colonial and postcolonial authorities to enshrine in law an "offense" of malign magic and thus to promote its negative antisocial image and sanction its organized persecution. This matter is discussed more fully below. 4. The reason for such legal invisibility in the British and Dutch areas of influence is closely linked to the predominance of Islam. Early Spanish cultural penetration cast witchcraft and sorcery as the work of the Devil, going some way toward explaining the higher social profile accorded the phenomena in the modern Catholic Philippines (see, e.g., the extracts from the Missions to Visayas, 1751-1765, in Blair and Robertson 1973:48, 113). Similarly, it may not be entirely unrelated that our best accounts of African witchcraft are from areas of Christian, not Muslim, influence. 5. In Africa the failure of British colonial law to distinguish between different facets of the same role, between curers and bewitchers, resulted in the prosecution of those accused of curing victims and therefore appeared to condone witchcraft itself (Orde-Browne 1935; Roberts 1935). In New Guinea, both colonial and postcolonial ordinances established a firm distinction between malign and harmless magic, thus separating a phenomenon which in many instances was conceptualized as indivisible into what many Africanist ethnographers would recognize as "witchcraft" and "sorcery." Similar confusion was avoided in colonial Southeast Asian law only by ignoring the subject altogether. 6. Although Turner (1964) deprecates the distinction, he fails to offer any practical advice on how the terminological problem might be resolved or how we might conceptualize different forms of mystical action. Barnes, J. A. 1962. African models in the New Guinea highlands. Man 62: 5-9. Baroja, J. C. 1964. The world of the witches. Chicago: University of Chicago Press. Barton, R. F. 1949. The Kalingas: Their institutions and custom law. Chicago: University of Chicago Press. . 1969. Ifugao ¡aw. Berkeley and Los Angeles: University of California Press. Blair, E. H., and A. Robertson, eds. 1973. The Philippine Islands, 1493-1898. Manila: Cachos Hermanos. Cedderoth, S. 1990. Javanese and Sasak folk beliefs: The changing role of diviners in two Indonesian villages. In Agrarian society in history: Essays in honour of Magnus Morner, ed. M. Lundahl and T. Svensson. London: Routledge & Kegan Paul. Colson, A. C. 1971. The prevention of illness in a Malay village: An analysis ofcon-

Introduction

23

cepts and behaviour. Developing Nations Monograph Series 2:1. WinstonSalem, N.C.: Wake Forest University, Overseas Research Center. Crawford, J. R. 1967. Witchcraft and sorcery in Rhodesia. London: Oxford University Press. Crick, M. R. 1976. Explorations in language and meaning: Toward a semantic anthropology. London: Malaby Press. Davis, R. B. 1984. Muang metaphysics: A study in Northern Thai myth and ritual. Bangkok: Pandora. Douglas, M., ed. 1970. Witchcraft confessions and accusations. ASA Monograph in Social Anthropology 9. London: Tavistock Publications. Ellen, R. F. 1976. The development of anthropology and colonial policy in the Netherlands, 1800-1960. Journal ofthe History ofthe Behavioural Sciences 12:303-324. . 1983. Social theory, ethnography and the understanding of practical Islam in South-East Asia. In Islam in South-East Asia, ed. M. B. Hooker, 50-91. Leiden: E. J. Brill. Endicott, K. M. 1970. An analysis ofMalay magic. Oxford: Clarendon Press. Evans-Pritchard, E. E. 1937. Witchcraft, oracles and magic among the Azande. Oxford: Clarendon Press. Favret-Saada, J. 1980. Deadly words: Witchcraft in theBocage. Cambridge: Cambridge University Press. Forth, G. 1989. Animals, witches, and wind: Eastern Indonesian variations on the "thunder complex." Anthropos 84:89-106. Fortune, R. F. 1932. Sorcerers ofDobu: The social anthropology oftheDobu islanders ofthe Western Pacific. London: Routledge & Kegan Paul. Fox, R. 1967. Kinship and marriage. An anthropologicalperspective. London: Penguin Books, Penguin Anthropological Library. Gimlette, J. D. 1929. Malay poisons and charm cures. London: J. and A. Churchill. Gimlette, J. D., and H. W. Thomson. 1971 [1939]. A dictionary of Malayan medicine. Kuala Lumpur: Oxford University Press. Gluckman, M. 1944. The logic of African science and witchcraft. Human Problems in British Central Africa 1:61-71. . 1967. The judicial process among the Barotse. In law and warfare: Studies in the anthropology of conflict (American Museum Sourcebooks in Anthropology), ed. P. Bohannan. New York: The Natural History Press. Golomb, L. 1985. An anthropology of curing in multiethnic Thailand. Urbana: University of Illinois Press. Gullick, J. M. 1988. W. W. Skeat and Malayan ethnography: An appreciation. Journal of the Malaysian Branch of the Royal Asiatic Society 61, pt. 1, 117-152. Hood Salleh. 1984. Introduction. In Malay magic, ed. W. W. Skeat. Singapore: Oxford University Press. Horton, R. 1967. African traditional thought and Western science. Africa 37:50-71; 155-187. Howe, L. E. A. 1984. Gods, people, spirits and witches: The Balinese system of personal definition. Bijdragen totde Taal, landmen Volkenkunde 140:193222.

24

ROY ELLEN

Keyes, C. F. 1980. Funerary rites and the Buddhist meaning of death: An interpretative text from Northern Thailand. Journal of the Siam Society 68 (1): 1-28. Kluckhohn, C. 1944. Navaho witchcraft. Harvard University, Mass.: Peabody Museum Papers 22 (2). Kohlbrugge, J. H. F. 1907. Blikkm in het zieleleven van den Javaan en zijner overheerschers. Leiden: E. J. Brill. Larner, C. 1981. Enemies ofGod: The witch-hunt in Scotland. London: Chatto & Windus. . 1984. Witchcraft and Belgian: the politics of popular belief. Oxford: Basil Blackwell. Lévi-Strauss, C. 1963. The sorcerer and his magic. In Structural anthropology, 161-180. New York: Basic Books. Lieban, R. W. 1967. Cebuam sorcery: Malign magic in the Philippines. Berkeley and Los Angeles: University of California Press. LuKrmann, T. M. 1989. Persuasions of the witch's craft: Bitual magic and witchcraft in present-day England. Oxford: Basil Blackwell. Macfarlane, A. 1970. Witchcraft in Tudor and Stuart England. London: Routledge & Kegan Paul. . 1970. Witchcraft in Tudor and Stuart Essex. In Witchcraft confessions and accusations, ed. M. Douglas, 81-103. ASA Monograph in Social Anthropology 9. London: Tavistock Publications. Mair, L. 1969. Witchcraft: An introductory survey. London: Weidenfeld & Nicolson. Malinowski, B. 1926. Crime and custom in savage society. London: Routledge & Kegan Paul. Marwick, M. G. 1965. Sorcery in its social setting: A study of the Northern BhodesianCewa. Manchester: Manchester University Press. . 1967. The study of witchcraft. In The craft